Fetal activity parameters classification from multidimensional Doppler signals

This paper shows preliminary feasibility to separate normal and pathological fetuses using a purposed built multi-transducer-multi-gate Doppler system and developed dedicated signal processing techniques via fetal activity parameters extraction. A dataset consisting of two groups of fetal signals (normal and pathological) has been established and provided by physicians. From fetal activity estimated parameters, an instantaneous Manning-like score, referred to as ultrasonic score was introduced and was used together with movements, heart rate and the associated parameters (variability, accelerations) in a classification process using Support Vector Machines (SVM) method. The influence of the fetal activity parameters and the performance of the SVM were evaluated using the computation of sensibility, specificity, percentage of support vectors and total classification accuracy. We showed the ability of our system to separate the data into two sets: normal fetuses and pathological fetuses and obtained an excellent matching with the clinical classification performed by physicians. (C) 2011 Elsevier Masson SAS. All rights reserved.